Why CalmStack doesn’t use streaks

A lot of apps want you to come back tomorrow. Then the day after that.

Published 4 September 2026 · Last evidence review 4 September 2026 · OASYS Editorial

A lot of apps want you to come back tomorrow.

Then the day after that.

Then the day after that.

Eventually, there is usually a number involved.

Seven days. Fourteen days. Thirty days. A little flame, a growing chain, some increasingly urgent indication that you are about to ruin something you have apparently spent weeks building.

Streaks are good at one thing: making continuity visible.

That can be useful. It can also change what the activity is for.

CalmStack doesn’t use streaks because the aim is not to keep you opening CalmStack. The aim is to make something useful available when you need it, and then let you get on with your day.

That sounds like a small design choice.

It isn’t quite.

A streak can become the thing you are trying to protect

There is nothing inherently wrong with streaks.

They can make progress visible. They can provide a prompt to return. They can help turn an occasional behaviour into a routine.

Research on streaks suggests there is another effect worth paying attention to.

Silverman and Barasch studied how people respond when a logged behavioural streak is intact or broken. Across seven studies, they found that the streak itself could influence subsequent behaviour. People did not simply respond to what they had been doing. They also responded to the state of the streak.

That distinction matters.

Once a streak becomes meaningful, maintaining the streak can become a goal in its own right.

You can end up protecting the number rather than doing the thing the number was originally supposed to represent.

That does not make the number useless. It means the number has acquired a job of its own.

Engagement is not the same as effectiveness

This is where app design gets slightly slippery.

If someone opens an app every day, that is easy to count.

Whether opening the app was useful is harder.

A 2025 randomised study involving more than 60,000 pupils in Peru provides an interesting example. Messages that highlighted streaks increased platform use and were associated with improved mathematics achievement compared with a control condition. That is a useful result for anyone designing a learning platform.

But notice what the finding does not mean.

It does not mean streaks are universally beneficial. It does not mean more frequent use is always better. And it certainly does not mean that a streak is evidence that a person is getting better at whatever the app is designed to help with.

The same distinction appears in broader research on gamification.

A 2024 meta-analysis of 36 trials involving more than 10,000 participants found that gamified digital health interventions produced some small improvements in outcomes such as physical activity and body composition. The effects were not universal, and the authors did not establish that gamification is a general solution for behaviour change.

So there is a perfectly reasonable case for using game mechanics.

CalmStack is simply trying to optimise for something different.

What are we actually counting?

Imagine two people.

One opens CalmStack every morning for 90 days. They rarely need anything from it, but they like maintaining the streak.

Another person opens it three times in a month.

On the first occasion, they notice that they are unusually wound up after a difficult meeting. They learn something about what is happening, try a brief exercise, and carry on.

On the second occasion, they notice a similar pattern.

On the third, they recognise it earlier.

Which person is doing better?

The app cannot answer that from the number of openings.

And that is the problem with treating engagement as the main outcome.

For some products, frequency genuinely is part of the objective. Language learning, revision and physical activity can all involve behaviours where repeated practice is central to the outcome.

Emotional regulation is less tidy.

Sometimes practice matters.

Sometimes noticing matters.

Sometimes learning something once is enough to change what you do later.

Sometimes the most useful outcome is that you opened the app, did one thing, and closed it again.

There is no particularly good reason to turn that into a competition with yesterday.

The number can change the relationship with the tool

There is also a motivational question here.

Self-determination research has long distinguished between doing something because it is personally meaningful and doing it because of an external reward or pressure. Digital products do not operate outside that distinction.

A streak is a form of feedback.

Depending on the person and the context, that feedback can be motivating. It can also introduce a new reason to continue that has little to do with the original purpose.

That is not a moral judgement about gamification.

It is a design consideration.

If CalmStack tells you that you have maintained a 42-day emotional-regulation streak, what exactly have you achieved?

The honest answer is: you have used CalmStack on 42 consecutive days.

That is all the number can establish.

It cannot tell you whether those sessions were useful.

It cannot tell you whether you needed them.

It cannot tell you whether you handled a difficult moment differently afterwards.

And it cannot tell you whether using CalmStack less often would actually have been better.

CalmStack is designed around usefulness, not attendance

That leads to a fairly simple principle.

CalmStack does not want to maximise app opens.

It wants the learning and practice to be useful when they are relevant.

That is why the product is structured around learning, noticing, brief practice and reflection rather than a scoreboard.

The distinction is important because it affects what the product measures and what it chooses not to measure.

A Check-in can tell you what you reported at a particular moment.

Repeated Check-ins can sometimes help you notice patterns.

A completed exercise can tell you that you completed the exercise.

None of those things needs to become a score for how well you are doing at being a person.

There is enough performance management in ordinary life already.

What this does not mean

It does not mean streaks are bad.

It does not mean gamification is bad.

It does not mean daily practice is unnecessary.

And it does not mean CalmStack has somehow discovered a superior universal theory of behaviour change.

The evidence does not support those claims.

The decision is narrower.

For this product, a streak would encourage us to make continuity visible even when continuity is not necessarily the outcome we care most about.

So we leave it out.

That also gives us a useful test for future features.

If a feature makes CalmStack more likely to be opened but less clear about what the user is actually getting from it, we should probably stop and ask what we are optimising.

That question matters more than whether the feature looks good on a dashboard.

The useful question

You do not need to maintain a perfect record for a tool to be useful.

You can learn something on Tuesday and use it on Friday.

You can notice the same pattern three times over a month.

You can use a brief exercise once and decide it is not for you.

You can take a Check-in because you are curious about what is happening, not because you need to keep a number alive.

There is no broken chain to repair.

There is just the next useful moment.

Try this

If the number disappeared tomorrow, ask:

Would I still want to do this because it is useful?

If the answer is yes, the behaviour probably has a reason to exist beyond the streak.

If the answer is no, that is useful information too.

Emotional regulation Everyday moments At work CalmStack